ESA UNCLASSIFIED - For Official Use

The BIOMASS Mission

Klaus Scipal

24/01/2019

1. ESA’s 7th Earth Explorer Mission

2. An interferometric, polarimetric P-band SAR

3. To be deployed in Space in 2021

ESA’s 7th Earth Explorer to be deployed in 2022

An interferometric, polarimetric P-band SAR

Designed to observe forest height and biomass

The BIOMASS Mission

ESA’s satellite programs

SO

URCES

34.1±1.7 GtCO2/yr 91%

3.5±1.8 GtCO2/yr 9%

16.4±0.4 GtCO2/yr

44%

Global Carbon Project, 2015

11.5±3.1 GtCO2/yr

31%

9.7±1.8 GtCO2/yr

26%

PARTIT

ION

ING

Fate of Anthropogenic CO2 Emissions (2005-2015)

What information do we need?

1. We need estimates of forest biomass, height and disturbances

2. The crucial information need is in the tropics:

deforestation (~95% of the Land Use Change flux)

regrowth (~50% of the global biomass sink)

3. Biomass measurements are needed where the changes occur and at the effective

scale of change: 4 hectares

4. Measurements are needed wall-to-wall

5. A biomass accuracy of 20% at 4 hectares, comparable to ground-based

observations

6. Detection of deforestation at 0.25 ha

7. Repeated measurements over multiple years to identify deforestation and

regrowth

How to measure biomass?

Tree allometry links biomass to

𝐴𝐺𝐵 = 𝜌 ∙ 𝐷2 ∙ 𝐻

DiameterHeightWood

density

Lopè

Mabounie Rabi

Mondah

HH+VV HV HH-VV

Tropical Forest as seen by DLR‘s P-band F-SAR

LN

L-1

0 (

MA

R1

)

LVIS P - HH L - HH P - HV L - HV

LN

L-0

5(C

OL

2)

LVIS P - HH L - HH P - HV L - HV

LN

L-0

8(O

KO

2)

LVIS P - HH L - HH P - HV L - HV

Information content of P-, L- band SAR and lidar

Courtesy: M. Pardini - DLR

PolInSAR(Polarimetric SAR Interferometry)

x

y

z

o

PolSAR(SAR Polarimetry)

x

y

z

o

TomoSAR(SAR Tomography)

x

y

z

o

SAR can deliver 3 independent types of information related to biomass

Biomass Mission Concept

Single satellite, operated in a polar

sun-synchronous orbit

Full polarimetric P-band (435 MHz)

Synthetic Aperture Radar with 6 MHz

bandwidth

Two mission phases: Tomography (year

1), Interferometry (year 2-5)

Multi-repeat pass interferometry (3

passes in nominal operations) with a 3

days repeat cycle

Global coverage in ~7 months (228

days) on asc. and des. passes

5 years lifetime

PolInSAR

(Polarimetric SAR

Interferometry)

x

y

z

o

PolSAR

(SAR Polarimetry)

x

y

z

o

TomoSAR

(SAR

Tomography)

x

y

z

o

Biomass Mission Requirements

Key Parameters

Sensitivity (NESZ) ≤ -27 dB

Total Ambiguity Ratio ≤ -18 dB

SLC resolution ≤ 60m x 8m

Dynamic Range 35 dB

Radiometric Stability ≤ 0.5 dB

Radiometric Bias ≤ 0.3 dB

Crosstalk ≤ -30 dB

✔

✔

✔

✔

✔

✔

✔

Tim

e [d

ays]

Across-track distance at the equator [km] 0910 150

Major Cycle

0

36

72

108

144

180

216

Satellite repositioning maneuver

Satellite repositioning maneuver

Satellite repositioning maneuver

Satellite repositioning maneuver

Satellite repositioning maneuver

Satellite repositioning maneuver

Global Coverage Strategy

Global Revisit Pattern

Coverage

1. Systematic Acquisitions for forested land (red area)

2. Global coverage in 7.5 months (INT phase) and 14 months (TOM phase).

3. Best effort acquisitions for non forested areas (yellow + ocean/sea ice ROIs)

4. Acquisition mask restricted by US Space Objects Tracking Radar (SOTR)

(Red = Primary objective coverage mask, Yellow = Secondary objective coverage mask)

Lopè

Mabounie Rabi

Mondah

HH+VV HV HH-VV

AfriSAR sites as seen by P-band by DLR‘s F-SAR

P-band SAR measures biomass and quantifies landscape dynamics

Yellowstone Park, 2003

Biomass0

A week after burn- 27dB

15 years after burn- 19dB

60-80 years after burn- 12dB

HV Backscatter

P-band SAR image (HH, VV, HV)

Global consistency in the biomass – P-band backscatter relationship

The simplest inversion: Similar power-law relationships between backscatter

and biomass are found for all forests where we have data

log 𝐴𝐺𝐵 = 𝑎 + 𝑏 ∙ 𝛾0𝐻𝑉

Strong variations in backscatter signal

© A. Monteith

Key challenge - how to estimate biomass

1. Loss of sensitivity in the high biomass range even at P-band.

2. Environmental nuisance factors (vegetation and soil water changes,

freeze/thaw, wind, …).

SAR tomography, a new concept to explore 3D forest structure

Generates images of different forest layers from multi-orbit SAR images

10.90.80.70.60.50.40.30.20.10

Normalisedbackscatter

intensity

TomographicProcessing

40

30

20

10

0

Height (m)

Guyaflux tower(Tropiscat experiment)

Tomographic imaging in Paracou

D. Ho Tong Minh et al., “Relating P-band SAR tomography to tropical forest biomass”, TGRS, Feb. 2014.

Key challenge - how to estimate biomass

1. Loss of sensitivity in the high biomass range even at P-band.

2. Environmental nuisance factors (vegetation and soil water changes,

freeze/thaw, wind, …).

Biomass enables a new approach to tackle these problems: notch out the ground

PolInSAR(Polarimetric SAR Interferometry)

x

y

z

o

SAR can deliver 3 independent types of information related to biomass

Interferometric ground notching

Idea: cancel out ground scattering by taking the difference between two phase

calibrated SLC BIOMASS images

Principle: SLC = projection of modulated

target reflectivity along elevation

𝑠1 = න𝑃 𝑧 ∙ 𝑑𝑧

𝑠2 = න𝑃 𝑧 ∙ 𝑒𝑥𝑝 𝑗𝑘𝑧𝑧 ∙ 𝑑𝑧

Master:

Slave:

Ground notched image = Slave – Master

𝑑 = 𝑠𝟐 − 𝑠𝟏 = න𝑃 𝑧 ∙ 𝐻 𝑧 ∙ 𝑑𝑧

H(z) = Vertical Impulse Response Function (IRF)

252525

Ground rejection greatly improves correlation and sensitivity

AGB vs TropiSAR backscatter

volu

me

com

ponen

t

pow

er [

dB

]

|slc

(hv

)|2

[dB

]

Single image

6 i

mag

es

30m

pow

er

[dB

]

Ground truth AGB [T/Ha*100]

TomoSARGround notched

Ground truth AGB [T/Ha*100]Ground truth AGB [T/Ha*100]

Forest biomass

Above-ground biomass(tons/hectare)

Upper canopy height (meter)

Areas of forest clearing (hectare)

• 200 m resolution

• 1 map every 6 months

• global coverage of forested

areas

• accuracy of 20%, or 10 t ha–1

for biomass < 50 t ha–1

• 50 m resolution

• 1 map every 6 months

• global coverage of forested

areas

• 90% classification accuracy

What information will we get from Biomass

• 200 m resolution

• 1 map every 6 months

• global coverage of forested

areas

• accuracy of 20-30%

Forestheight Disturbances

Biomass will allow DEM production under dense tropical canopies

90m x 90m DEM feasible with ~ 2 m height accuracy from Biomass

TropiSAR data

Image courtesy of P. Dubois-Fernandez

Paracou

Optical Imagery (Google Earth)

P-band enhances subsurface imaging in arid zones

L-band SAR (ALOS)P-band SAR

Summary – BIOMASS a true Earth Explorer

1. BIOMASS implementation started in Nov. 2013. We just kicked-off Phase-C

(CDR planned in 2019). We are working towards a launch in 2022.

2. BIOMASS is the first P-band SAR and first radar tomographic space mission; it

is a true Earth Explorer with a lot of unknowns and exciting science for global

biomass mapping.

3. The new unique vision of Earth from Biomass will extend beyond forests and

into measurements of ice, sub-surface geomorphology in deserts, topography,

the ionosphere, ocean ...

1. ESA’s 7th Earth Explorer Mission

2. An interferometric, polarimetric P-band SAR

3. To be deployed in Space in 2021